Thema

#sandbox

Loots, Blogposts und verwandte Themen rund um diesen Tag. Folge dem Tag, damit passende Updates in deinem Orbit bleiben.

#sandbox
2Gezeigte Loots
7Gezeigte Artikel
8Verlinkte Nachbar-Tags
Anschluss-Themen

Wenn du tiefer einsteigen willst, helfen die benachbarten Tags beim Vergleichen und Querlesen.

#developer-tools#AI agents#coding-agents#vercel#ai-generated#automated#cloud#open-source
Loot

Mehr aus diesem Thema

Alle Loots entdecken
Give Any Model a Sandboxed Shell and File Workspace with OpenRouter

Give Any Model a Sandboxed Shell and File Workspace with OpenRouter

No votes yet
Text: AI-generated
#ai-generated#automated+7
AI-generated · Automatically published by LinkLoot. A practical OpenRouter beta for agent builders: let compatible models run commands in an isolated Linux container and move input or output files through the Files API. AI-generated: This Loot was created and published automatically by LinkLoot and was not substantively reviewed by a human editor. What it is OpenRouter now exposes openrouter:shell on its Responses and Messages APIs. A tool-calling model can request shell commands, receive stdout, stderr, and exit outcomes, and continue working from those results. The companion Files API lets you upload workspace files, attach them to a container, and retrieve files created by a run. Useful for Agent workflows that need to transform CSV, Markdown, PDFs, or other supported files. Prototypes that need repeatable server-side scripts without running commands on the users own machine. Multi-model applications that want a shared hosted tool surface across providers. Access and limits The shell and Files API are beta features on the global openrouter.ai endpoint. Shell is available through the Responses and Messages APIs, not Chat Completions. Containers have outbound networking disabled by default; an allowlist can be configured when a job genuinely needs egress. Commands are bounded by server-enforced time and output limits, and containers can sleep after five minutes of inactivity. Sandbox time is billed at $0.0001 per active second, with a 30-second minimum when a cold container starts. Files API storage has no separate charge, while workspace storage is limited to 10 GiB. Check the current documentation before relying on beta behavior or sending sensitive data. Start here Read the announcement for the workflow overview, then use the Shell and Files API documentation for request shapes, container policy, file retention, and endpoint restrictions. Treat model-generated commands as untrusted: keep network access allowlisted, avoid secrets in prompts or files, and review the commands your application permits. Sources OpenRouter announcement: Give any model a terminal and files OpenRouter Shell server tool documentation OpenRouter Files API documentation
LinkLoot access
Free
Provider costs
Unknown
Review open
0
Open loot
Run coding agents in disposable Linux VMs with Clawk

Run coding agents in disposable Linux VMs with Clawk

No votes yet
#coding-agents#sandbox+5
Clawk gives Claude Code, Codex, and shell-based coding agents a disposable, network-restricted Linux VM so they can install tools and run code without direct access to your host machine. What it is Clawk is an open-source agent sandbox for local development. You start it from a repository, then run a coding agent or shell inside a disposable Linux VM with restricted outbound networking and only the mounted project files available. Why it is useful Use it when an agent needs to install packages, run servers, execute generated code, or inspect unfamiliar dependencies, but you do not want that work happening directly on your laptop. The project is pre-1.0, so treat it as a practical experiment rather than a hardened enterprise boundary. Best fit Use case Why Clawk helps Caveat --- --- --- Agent coding sessions Gives the agent root inside a throwaway Linux guest Anything mounted or allowed on the network can still be exposed Risky dependency tests Lets packages run away from the host filesystem You still need normal code review and secret hygiene Multi-agent experiments Keeps destructive commands away from the main machine Pre-1.0 project with possible breaking changes Before you try it Check the supported platforms, read the security model, and start with a non-sensitive repository. Do not mount secrets or private data unless you are comfortable with the agent and allowed network destinations seeing them.
LinkLoot access
Free
Provider costs
Unknown
Review open
0
Open loot
Blog

Verwandte Artikel

Blog durchsuchen
Tools & Apps

Vercel Sandbox reaches four regions for lower-latency agent workloads

Vercel Sandbox is now available in four regions, adding Paris alongside three U.S. locations and introducing project-level failover settings…

Tools & Apps

Vercel moves Sandbox defaults to versioned managed images

Vercel Sandbox SDK v3 now defaults new sandboxes to open-source managed images, changing the base operating system, toolchain, and update mo…

Tools & Apps

Vercel Sandbox stops billing internet downloads for agent runs

Vercel Sandbox no longer counts package installs, repository clones, external artifacts, or dataset downloads toward Sandbox Data Transfer u…

Tools & Apps

Mount S3 in Vercel Sandbox before your agents copy every dataset

Vercel Sandbox now supports FUSE-based filesystems, so agent and developer workloads can mount S3, network filesystems, or custom storage dr…

Tools & Apps

SuperHQ Puts Coding Agents Inside Local microVM Sandboxes

SuperHQ is an early open source app for running AI coding agents in isolated local microVMs, with diff review and an auth gateway that keeps…

Tools & Apps

Runtime brings sandboxed coding agents to whole teams, not just individual developers

Runtime launched on Hacker News with a team-oriented agent runtime for Claude Code, Cursor, Codex, Gemini CLI, and similar tools, combining …

AI & Automation

OpenAI Agents SDK Adds Native Sandbox Execution for Production Agent Workflows

OpenAI's updated Agents SDK adds a model-native harness and native sandbox execution, giving teams a more controlled way to run long-running…